Security Engineer

Runable · Full-time · In person, Bangalore

What we're building

Runable helps people build, grow, and run their businesses with AI agents. The work spans products customers use, agents that act across tools, our own GPU inference, and payment integrations with financial partners around the world.

We use our own product to build and test interactive artifacts and designs. Our agents monitor security around the clock, investigate and fix production issues, and detect abuse and enforce bans. We experiment with models and how they behave. This is already part of how we work.

We question the starting assumptions. When the existing approach to remote desktops was not good enough, we built Tilt in Rust to rethink the noVNC stack and protocol. We want people who ask what could work better, then build it and prove it.

The role

Own the security of a product where agents use tools, run code, operate computers, and interact with payment integrations. You will work inside the engineering team, close to the code and the people using it. The result is stronger boundaries and safer customer outcomes, with evidence that the defenses work.

What you'll own

• Threat-model agent tools, sandbox execution, browser actions, identity, shared resources, and payment permissions. Find where untrusted content can become an action or cross a boundary.
• Review and improve authorization, tenant isolation, secrets handling, dependency security, and the software delivery path.
• Build on the agents already monitoring security and abuse around the clock. Improve detection, investigation, remediation, and the controls around automated enforcement.
• Investigate incidents and abuse, contain the immediate problem, fix the shared cause, and verify the fix in the running system.
• Work directly with engineers and customers on security reports. Turn findings into shipped improvements and useful evidence rather than a queue of recommendations.
• Make the safe path easy to use, and measure whether our exposure and time to detect and resolve problems actually improve.

What you'll bring

• Concrete security work: vulnerabilities you found and responsibly reported, systems you hardened, incidents you handled, or tools people use.
• The ability to read unfamiliar code, reproduce a failure, explain its impact, and ship or work through the fix.
• Strong foundations in web applications, authentication, authorization, operating systems, and cloud infrastructure.
• Curiosity about prompt injection, agent permissions, model behavior, and the risks of letting software take actions for people.
• Judgment about severity, customer harm, and tradeoffs. A certification can be a signal; demonstrated work is what we evaluate.

How we work

• Everyone at Runable writes code and ships, including design and GTM. Use agents to extend what you can build, and verify the work yourself.
• Own the outcome, from understanding the customer to shipping, operating, and measuring the result. Your job extends beyond the ticket or your title.
• Everyone can talk to customers. Go directly to the people using what you build, help them, and bring what you learn back into the product.
• AI agents are part of the team. Use them to research, build, review, test, and operate; your judgment owns the result. Build the automation that makes the next piece of work easier.
• Trust compounds. We trust people shipping code to merge to main when checks and evidence support it, and to ask for review when judgment or risk calls for another pair of eyes. Automated checks and approvals help us move; they do not transfer responsibility.
• Own the metrics and revenue impact of your work. Define the customer and business result, measure what changed after shipping, and keep improving it. Useful outcomes, sustainable revenue, and systems we can afford to run belong together.
• We work six days a week, in person in Bangalore. You have autonomy over how you organize and do the work. We care about outcomes and dependable collaboration, without managing every small decision.

What matters to us

Show us what you have built, how you think, and how ambitious you are. Degrees, employer names, and years of experience can be signals; none is a substitute for the work. We value curiosity, clear communication, good judgment, and the drive to take an unfamiliar problem all the way to a useful result.

Hiring process

1. A 15-minute introduction with Saksham, co-founder.
2. A work-day trial in our Bangalore office on a real problem with the team, so both sides can see what working together is like.

Apply

Email team@runable.com with links to work you are proud of, what you personally owned, and what changed because you built it. Code, products, research, writing, experiments, and customer outcomes all count. Tell us what you want to build next. A short, specific note is more useful than a polished resume.